Different methods for making windmill blades are known.
Thus it is known that windmill blades may be made by winding roving tapes or roving bundles around a core or mandrel. Methods for this are inter alia described in U.S. Pat. Nos. 4,242,160 and 4,381,960.
Such methods by winding have the disadvantage that after setting, the winded item will normally appear with the raw composite material surface as an external surface which is incompatible with many applications, e.g. for windmill blades. A satisfactory surface quality therefore presupposes a finishing treatment, e.g. by the bonding of shells made separately.
Another drawback by this method is that the winding normally presupposes use of a mandrel with a certain strength which therefore is desired to be reused. In these cases, the method may only be used with items having a geometry allowing removal of the mandrel, which means that the dimensions of the internal cross-section of the cavity at a given distance from the end from which the mandrel it drawn out are not to exceed the dimensions of any of the cross-sections situated between the position in question and the end, and that some tapering in the mold will normally be required in practice. Such a method may thus not be used for e.g. tanks or whole windmill blades.
It is also prior art that windmill blades may be made by a method where a blade is usually made with two half-shells which are joined at leading and trailing edges by bonding. The half-shells are usually supported inside the blade cavity by one or more beams, which are also joined to the half-shells by bonding, where the beams e.g. may be made in U- or I-shape so that the flanges of these beams form contact surfaces with the half-shells, or where the beams e.g. may be made by winding so that a part of the external surface of the winded beam forms contact surfaces towards the half-shells. The half-shells may e.g. be made of dry fiber materials which are supplied resin by manual laying, vacuum injection or the like, or they may be made of prepeg, where the fiber materials are impregnated in advance with resin which is brought to set by the action of heat, UV-irradiation, or similar. In other embodiments, beams and/or half-shells are made of thermoplastic, e.g. by using fiber materials that are combinations of temperature resisting fiber materials and thermoplastic, and where the fiber material after laying is brought to a temperature where the thermoplastic material melts, thereby acting as resin in the finished laminate.
However, it is a problem with this method that it may be difficult to ensure a satisfying quality of the glue joints established in the interior of the structure for the mutual joining of the half-shells and for joining possible beams with the half-shells. This is partly due to fundamental problems regarding material technology, partly to more specific manufacturing problems.
The fundamental problems regarding material technology may summarizingly be described as consequences of the impossibility of having the same material properties in the glue as in the rest of the blade. The reason for this is that the general material properties in the blade shells and the laminates of the possible beams are determined by the fiber reinforcement, which normally has rigidity several orders of magnitude higher than that of the resin, whereby the properties of the resin has minimal significance for the rigidity of the finished laminate. Conversely, the glue is normally made as pure resins (which may consist of other plastics than those used in the laminates) or as mixtures of resins and fillers but without fiber reinforcement. The result is that the elastic modulus of the glue typically deviates an order of magnitude, often several orders of magnitude, from the parts joined with the glue. To this comes that glue materials are often brittle and may therefore be vulnerable to local moments tending to open the glue joint, so-called peeling. Such local moments will particularly occur by very large loads on the blade, where non-linear effects may imply the blade cross-section changing its shape. By virtue of glue materials normally having relatively brittle properties, there may be the subsequent danger that cracks in glue joints propagate far beyond the area in which the original overloads have occurred.
Among the manufacturing problems, one of the essential is that the glue joints are provided at leading and trailing edge and between beam and shell, so that a glue joint is established on the unprepared surface at the inner side of the shell laminate. The problem of this joint is that the glue surface may only be defined within a certain large range of tolerance. To this comes that in the case of the trailing and leading edge bond the shell laminate has to be reduced towards the edge of the shell when, as e.g. in the case of windmill blades, the case is half-shells where the edges are abutting mutually inclining in order that the glue joint can have nearly uniform thickness. This reduction may not always be provided with the necessary tolerances why a real adaptation will require working of the assembly faces, which in turn will imply a large rise in the costs. Another problem is that the deformations arising in the blade shells in connection with small variations in the manufacturing process can give a varying gap inside the cavity of the item so that it may be difficult to ensure a complete filling of glue of the interspace between beam and shell. All these problems with tolerances have the result that glue joints generally may have varying cross-sections and fillings which in turn implies a risk of considerable stress concentrations in the glue and the adjoining blade shells and beams. Furthermore, it is a problem that most glue materials presupposes that the surfaces to be bonded are ground in advance with the associated problems of maintaining the necessary tolerances. Finally, the glue joints are usually difficult to inspect visually as well as they are difficult to inspect by NDT methods (non-destructive testing) due to the tapering laminate and the irregular geometry of the item.
It is also a problem with methods based on bonding individual parts of blades that even though individual sections of the blades may be produced in closed processes with small or no environmental loads, this is usually not the case with the bonding itself. Here, workers will usually be exposed to grind dust from dry grinding, partly because it is unfavorable to the subsequent gluing process to perform wet grinding and partly because they are exposed to contact with and/or vapors from the glue material itself, implying need for personal protective means.